144 



relationship, will require a considerable amount of work and much critical sense 

 on account of the great variation within a series of structures. 



Membranipora L. 

 Biflustra d'Orb (p. p.). Busk, Smitt. 

 Nichtina Canu. ' 



The zocecia, the aperture of which is to a greater or smaller extent surrounded 

 by a granular or denticulate, cryptocyst margin, have 2 spines at most, which 

 are situated in the two proximal corners. On each side of the distal wall one 

 multlporous rosette-plate or a series of uniporous or partly multiporous; each 

 lateral wall with 2 — 4 multiporous plates. No avicularia; no ooecia. 



In M. membranacea the cryptocyst appears only as an extremely narrow mar- 

 ginal portion, while in other species it attains not only a greater breadth, but 

 also forms a large, proximal expansion, often terminating in a larger or smaller 

 process. This is most strongly developed in M. denticulata (danica) v. scutata^, in 

 which it almost reaches the operculum as a free, quadrangular lamina. It is less 

 developed in M. delicatula Busk. The cryptocyst attains its highest develop- 

 ment in M. oblonga Busk', which represents a Micropora-Uke development of 

 the genus. The two spines, which constantly appear in M. membranacea and in 

 M. tubercalata and which in a number of species are more or less tuberculiform, 

 attain their highest development in the latter species, in which they often coal- 

 esce into a single very large tubercle, and a similar coalescence takes place in 

 a number of zocecia of M. Lacroixi And* (non Busk, nee Hincks), figured by 

 Savigny, the spines of which generally seem to have a triangular transverse 

 section. In M. Savarti and M. denticulata these spines are not constant, and in a 

 series of undescribed forms, which must h^ referred to other species, they are 

 altogether absent. In my description of M. membranacea in »Zoologia Danica« I 

 have already called attention to the great variation in the rosette-plates of the 

 distal wall in this species. On either side there may be sometimes a large multi- 

 porous rosette-plate, sometimes a series of smaller, uniporous or partly multi- 

 porous ones. In all the forms examined by me multiporous rosette-plates are 

 constantly found on the lateral walls. 



As in all the numerous forms, I have had the opportunity of examining, no 

 ooecia were found, which however are said to be present in the species from 



• 11 a, p. 380. ^ 55, p. 54. ' 7, p. 34. * While the species of Savigny is furnished with two spines 

 which in some zooecia are coalesced into a semiglobose tubercle the species, which Hincks calls 

 Memb. Lacroixii, has a greater or lesser number of small triangular hollows (kenozooecia) between 

 tlie zooecia. 



